The nuclear matrix protein NuMA ( Nuclear Mitotic Apparatus) is present at greater than tenfold higher levels in malignant tissues relative to normal tissues. Scientists at Matritech have shown that interference with NuMA synthesis using antisense oligonucleotide technology dramatically slows the growth of cancer cells both in soft agar and in liquid culture. We believe that the combination of its overexpression in cancerous tissue and its essential role for growth of cancer cells make NuMA a very good anti- cancer therapeutic target. NuMA has a structure similar to that of the intermediate filament proteins such as cytokeratins and lamins. This motif indicates that it may be capable of both self association as well as binding other nuclear proteins. We propose to use the two hybrid cloning system to set up a high throughput cell based assay to discover small organic compounds which inhibit the self association of NuMA. We also intend to develop a biochemical assay measuring recombinant human NuMA self association which will be used to verify on a biochemical level that the compounds discovered in the two hybrid assay are antagonists of this process. PROPOSED COMMERCIAL APPLICATION: Cancer chemotherapy is compromised by the dose limiting side effects observed in treatment. The nuclear matrix filament protein NuMA has been shown to be overexpressed in cancer and to be necessary for the growth of cancer cells. We are proposing to discover compounds which inhibit NuMA polymerization using both a biochemical and a cell based assay.